Re: Strange trace files generated in RDBMS 19.6.

Ahhh.... I did not catch that it was a lightweight thread. That's one of the newer features that I don't see a need for on Linux/Unix, so I have not configured.

Perhaps your errors might clear up with the traditional processes?

Just a thought...

Rich

On Fri, Jun 5, 2020 at 1:39 AM Nilo Segura <nilosegura_at_gmail.com> wrote:

> Hello,
>
> Unfortunately, there is no trace of any abnormal behaviour in the
> alert.log file. I have also sampled v$session during ~60m and could
> not find a reference to the PROCESS value that would match the trace
> file names, so either I'm very unlucky or those process die before
> they register into v$session. All those processes terminate
> themselves, so it is not a background process that decides to kill it.
>
> Process termination requested for pid 10169_10297_LT [source = rdbms],
> [info = 2] [request issued by pid: 10169, uid: 1286]
>
>
> BTW, the lowercase in these trace files seems to match Bug 27681633 -
> LIGHTWEIGHT THREAD TRACEFILE DOES NOT SHOW CORRECT PROCESS NAME
>
> regards.
>
> On Thu, Jun 4, 2020 at 9:42 PM Rich J <rich242j_at_gmail.com> wrote:
> >
> > Hi Nilo,
> >
> > I'm also running 19.6, but on the slightly-different OL7.7 with UEK.
> While my trace files have aged out since the last time that happened, I
> still see the corresponding entries in the alert log. In my case, the
> process terminations were for instance shutdown and job queue process slave
> termination. Perhaps you might find more info in your alert log?
> >
> > Rich
> >
> > On Thu, Jun 4, 2020 at 4:08 AM Nilo Segura <nilosegura_at_gmail.com> wrote:
> >>
> >> Hello,
> >>
> >> We have a 19.6 db running on RH 7.x . We see that the db generates an
> >> unusual amount of trace files for no apparent reason.
> >>
> >> Jun 4 10:01 aisbip2_ora_8966_9012_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8982_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_9011_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_9009_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_9008_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_9010_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_9000_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_9007_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8997_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8991_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8992_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8990_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8989_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8988_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8986_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8987_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8985_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8984_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8981_lt.trc
> >> Jun 4 10:01 aisbip2_ora_8966_8983_lt.trc
> >>
> >> In each file, we see something like
> >>
> >> Dump file
> /ORA/dbs0a/AISBIP_AISBI_RAC52/diag/rdbms/aisbip_aisbi_rac52/AISBIP2/trace/aisbip2_ora_8966_8983_lt.trc
> >> Build label: RDBMS_19.3.0.0.0DBRU_LINUX.X64_190417Oracle Database
> >> 19c Enterprise Edition Release 19.0.0.0.0 - Production
> >> Version 19.6.0.0.0
> >> Build label: RDBMS_19.3.0.0.0DBRU_LINUX.X64_190417
> >> ORACLE_HOME: xxxxxxxxxxx
> >> System name: Linux
> >> Node name: xxxxxxxxxx
> >> Release: 3.10.0-1062.1.2.el7.x86_64
> >> Version: #1 SMP Mon Sep 16 14:19:51 EDT 2019
> >> Machine: x86_64
> >> Instance name: AISBIP2
> >> Redo thread mounted by this instance: 2
> >> Oracle process number: 0
> >> Unix process pid: 8966, thread id: 8983, image: _LT
> >>
> >>
> >> *** 2020-06-04T10:01:28.718971+02:00
> >> Process termination requested for pid 8966_8983_LT [source = rdbms],
> >> [info = 2] [request issued by pid: 8966, uid: 1286]
> >>
> >> And that's all... There are many many files (one per process/thread)
> >> generated there, and DIAG filesystem becomes unusable.. these trace
> >> files are generated almost constantly... No idea if this correspond
> >> to a real problem (but no users are complaining about killed
> >> sessions). It could be another useless trace file generated by the DB.
> >> We have only seen this in this DB after the upgrade to 19.x . We are
> >> not using the threaded architecture (threaded_execution=false)
> >>
> >> These trace files do not have an associated .trm file... only .trc.
> >> Besides, the filename convention is different... in the "normal"
> >> trace files, the instance name is in uppercase (AISBIP2), but here is
> >> lowercase (aisbip2).
> >>
> >> Anybody seen a similar behavior ?
> >>
> >> regards.
> >>
> >> Nilo Segura
> >> Database Support
> >> CERN - Geneva
> >> Switzerland
> >> --
> >> http://www.freelists.org/webpage/oracle-l
> >>
> >>
>
>
> --
> Nilo Segura
> Oracle Support - IT/DB
> CERN - Geneva
> Switzerland
>

--
http://www.freelists.org/webpage/oracle-l